The macro buttons went wrong at the end of a transmission, both faults in
DigitalEngineSender:
- The engine reports the transmitter drop on its own thread.
WhenTransmitChanged took the state lock, decided the message had ended,
released the lock, and only then called Buffer.Ended(), which clears
what has gone to the engine. A macro pressed on the last character got
through StartAsync in that gap and had already flushed its own text into
Sent, so Ended() wiped the new text off the pane while the engine
transmitted it. Ended() is now called inside the same lock.
- The pane only started again when the engine reported a drop. Two macros
in a row keep the transmitter up, so that report never came and Sent
grew with every press. Everything in Sent is locked, because it is in
the engine and cannot be taken back, so the whole pane became
read-only. TypeAhead.Started() drops the last message's sent text and
keeps what was typed ahead, and StartAsync calls it whenever it keys a
new transmission.
The rest of this commit is the digital transmit work these fixes sit on:
the pane as one coloured box, the sender's three keying states, the
type-ahead feeder paced by the clock with the engine's count as a brake,
{RX} flushing what is left in one piece, and the entry window's function
keys reading the digital macros on a digital mode.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PdAYHcdRktqKry7nk414TU
The third probe run puts numbers on both. TxBufLen lags by 100 to 150 ms, about
one character at 45.45 baud: twenty-one characters pushed at 2496 ms read 0 at
2497, 2547 and 2597, then 26 at 2647. It moves once per symbol transmitted, so
readings 50 ms apart repeat.
It read 26 for 21 characters because it counts Baudot symbols, and the message
carried two digits: a shift to figures and a shift back each time, plus one at
the start. That is the case the cap was needed for and I had not seen. A contest
exchange is mostly digits, so the engine always has more to transmit than the
clock thinks, and the clock alone would run ahead of it on every QSO.
Slack of three characters covers both, so nothing changes in the pump. The
count reaching 0 does not cut the end of a message off either: it emptied 350 ms
before the last character was decoded back and the engine held the transmitter
up for another 900 ms.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RtspmWmS7f8kUvcyaHpRWZ
The second probe run read TxBufLen as 0 one millisecond and again fifty
milliseconds after twenty-one characters had been pushed, while those characters
were already going out: the first of them was decoded back off the air 420 ms
later. The number is right when it is a second old and wrong when it is fresh,
so it runs behind the engine.
A pump that fed on it would have handed over the whole message in half a second
and put all of it beyond reach, which is worse than the clock it replaced. So
the clock is the pace again, and the count is a check on it: the engine is never
given more than Lead + Slack characters however fast the clock says to feed, and
a message ends when the count and the clock estimate both say it has. Characters
fed since the last answer are added to it, so a stale answer cannot be spent
twice.
The engine is set to Character out: ABCD with no space after it went out at
once. Word out would have held it, so the holding case stays, but it now needs
the count to be neither going down nor being added to, since an engine kept
exactly at the cap has a count that does not move either.
The probe writes down every reading for two seconds after a push and does not
believe an empty one in that window, which is what the run needed to measure the
lag and did not do.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RtspmWmS7f8kUvcyaHpRWZ
The engine probe answered two questions on the air, and they point opposite
ways.
TxBufLen is the number of characters left to transmit. Seventeen characters
pushed, five decoded back, and it read 12, counting down to 0 as the message
went out. So the pump no longer counts character times off the clock: it asks
the engine, feeds while the answer is under Lead, and asks again. The baud rate
is now only the fallback for an engine that will not answer.
A backspace is not an edit. Pushed in as a character it made the count go up by
four and the text went out unchanged, so text the engine has been given cannot
be taken back, only aborted. EngineTypeAhead was built on the opposite belief
and is gone, along with the setting that chose it and the interface that existed
to switch between the two.
A count that stops going down means the engine is holding what it has: MMTTY set
to Word out keeps a word until the space after it. The pump feeds one character
per look while that lasts, so the space arrives and the word goes out rather
than the message sitting there.
The probe kept two bugs of its own that this run showed: it read the count a
millisecond after pushing, saw 0 and called the message finished, and it started
a step while the previous unkey was still in flight, so that step ran with the
engine down. It now needs two empty answers in a row and waits for the engine to
say it has stopped. It also asks what {RX} does to a word the engine is holding,
which decides whether a macro without a trailing space loses its last word.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RtspmWmS7f8kUvcyaHpRWZ
Nothing was ever transmitted. SetMmttyPTT does not start a transmission: N1MM
calls it with 1 to stop once the buffer is empty, which is its XmitOff, and with
0 to stop now, which is its AbortXmit. A transmission starts by setting the
control's PTT property, in XmitOn.
So the bridge learns `key <0|1>` for that property, and MmttyEngine now keys
with it, ends a message with SetMmttyPTT(1) so the buffer still goes out, and
aborts with SetMmttyPTT(0). This is the digital {TX} in the entry window and the
digital window as well as the probe: none of them could key the engine before.
The probe checks that the engine keys before it measures anything, and stops
with a plain statement if it does not, rather than reporting numbers from an
engine sitting still. It also asks a new question: whether the engine holds a
word until the space after it, which is MMTTY's Way to send. Received characters
are marked as noise while the engine is not transmitting, since a machine with a
sound card decodes the band all the way through the run.
The first run on a real engine says MMTTY 1.70 connects, the control answers
TxBufLen and refuses NotAProperty with DISP_E_UNKNOWNNAME. What TxBufLen counts
is still open: read while nothing was transmitting it rose over time and rose by
four after four backspaces, which is not what characters-left would do.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RtspmWmS7f8kUvcyaHpRWZ
MMTTY has a type-ahead buffer of its own: characters go into it, a backspace
takes back one it has not transmitted, and TxBufLen says how many are left. Used
that way it paces itself, so there is no gap between characters to tune and no
baud rate to keep in step with the engine.
EngineTypeAhead does that, and Config > Digital picks between it and the pump
that is there now. Off is still the default: three things it rests on have never
been seen with a real engine.
tools/Nonemm.EngineProbe asks the engine those three questions and writes the
answers to a file. It starts MMTTY through the bridge, pushes a message, polls
TxBufLen while it goes out, backspaces over text that has and has not been
transmitted, and logs what came back on the receive side and when.
The bridge learns one verb for it: `buffer` reads TxBufLen and answers with the
count, or -1 when the control will not say. A property the control does not know
is a log line rather than an error, since it stops nothing.
TypeAhead and EngineTypeAhead share the TransmitBuffer interface, which is what
the digital window now works through, so the window does not know which one it
has.
docs/unfinished.md states what each buffer assumes and how to run the probe.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RtspmWmS7f8kUvcyaHpRWZ
The pump handed the engine one character and then waited a character time
before the next, on purpose, so it stayed behind the engine rather than ahead.
That leaves the engine with an empty buffer between every two characters of a
message, and an engine with an empty buffer transmits idle rather than waiting.
The idle is added to how long the message takes, which in a contest is time
paid for nothing.
The pump now keeps two characters in the engine: one being transmitted and one
behind it, so the engine never runs dry. The clock at the baud rate says when
the engine has room for the next one, and the engine reporting that it has
stopped transmitting sets the estimate back to an empty buffer.
The last two characters are now beyond reach rather than the last one. Everything
before them can still be rewritten, which is what the pane is for.
A message is finished when the buffer is empty and the engine is estimated to
have transmitted what it holds, so {END} and {RX} no longer run while the engine
still has the last characters of the message.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RtspmWmS7f8kUvcyaHpRWZ
The pump is paced off the clock. XMMT.ocx has a TxBufLen property and answers
when it is asked, so the bridge could report what the engine still holds and
make the red text exact, but the engine does not start on this machine and the
property could not be watched while a message was going out.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
The transmit pane is now the type-ahead buffer. What has gone to the engine is
in red beside the box and cannot be reached; what is still to go is in the box
and can be typed over, added to or deleted while the engine works through it.
A macro fills the pane rather than the engine, so the operator can correct a
call the message is still sending.
The cursor holds the pump: nothing behind it goes out, so an engine that
catches up with the operator idles instead of transmitting half a word. With
the box out of focus there is no cursor and everything goes.
Escape and the RX button drop what has not gone out and stop the engine where
it is. Enter puts in the carriage return the engine takes as a new line. The
pane starts empty for the next message.
The baud rate the pump counts by is in the digital setup window, beside the
alignment frequency. It has to match the engine: 45.45 unless the contest says
otherwise.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
A digital engine takes a whole message and transmits it at 45.45 baud, which
takes several seconds. Once it has the message nothing can be changed, so an
operator who sees the wrong call going out has to stop the transmission and
start again.
TypeAhead keeps the message instead and feeds the engine one character at a
time. What has not gone out yet can be rewritten, added to or deleted. Cursor
is how much of it may go: the pump stops there and the engine idles on the air,
which is the diddle a RTTY engine sends between characters anyway.
The pump counts character times off the clock at the baud rate rather than
asking the engine what it has left, so it runs a little behind: a gap between
two characters is idle on the air and costs nothing, while feeding faster than
the engine transmits would put text out of reach again.
Every digital message goes through it — the function keys, ESM, the QTC window
and the digital window's own buttons — because they all send through
DigitalEngineSender. Finished is now raised when the buffer runs dry, which is
when the message has really gone, so what stands after {END} runs then and {RX}
drops the transmitter at the right moment. The transmitter dropping only counts
as the end when nothing is left to send: an engine that keys itself off what it
is given drops between two characters as well.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Sending a text file is in, so it comes off the list. AutoTRXUpdate goes on it:
the title bar carries the offset frequency now, but the entry window and the
log still record the dial frequency.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
The window now carries what N1MM's does. Across the top: the TX indicator, the
Letters/Figs box showing the word under the pointer in the other Baudot case,
and the MouseOver box showing the word itself. Down the left of the receive
pane is the pause bar, green while text prints and yellow while it is held.
Under the panes are the message buttons and the row of engine controls, in
N1MM's order.
Both of N1MM's window scroll types are here. The scrolling pane adds lines at
the bottom; the non-scrolling pane keeps text where it was printed, underlines
the line being printed to and walks it down the window, so a call does not move
as the pointer reaches it.
A callsign is checked before it is coloured: anything shaped like a call under
the generic routine, only calls in the files that are ticked under the
resources routine, and under the combination everything, with a call the files
do not hold marked in yellow. The garbage text search finds a Master.scp call
inside a run of noise.
The mouse and the keys are N1MM's. Shift+click converts the case before the
word is transferred, Ctrl+click puts it in the box the cursor is in, Alt+click
puts a call on the stack, a double click replaces the call in the box. The
receive pane and the grab list have their right click menus. Alt+T and Ctrl+K
turn the transmitter on with the cursor in the transmit pane, Alt+G grabs, Esc
aborts.
New from the menus: hover mode, the callsign pause routines, sending a text
file, a date-stamped copy of the receive pane in ExportFiles, clearing the grab
list on a CQ or a QSY, the ham default and AFC off on a run to S&P change, AFC
on with CQ, and raising the window with the entry window it belongs to. The
grab list holds ten calls, as N1MM's does.
The setup window is N1MM's dialog: a General / MMTTY Setup tab, an Engine tab
for the Wine and bridge paths, and a Message Setup tab with the buttons laid
out as they sit in the window, a caption and text box for the one being edited,
the macro list, the count of buttons to show, and import and save of a .mc
file. Right-clicking a message button opens that tab, so the free-form macro
dialog is gone.
The title carries the dial frequency with the mark tone added to it, or the
dial frequency alone when the radio already reads out what it transmits.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
The settings N1MM keeps for its digital interface: how many message buttons to
show, which of the two window scroll types to use, hover mode, the callsign
pause routines, what a callsign is checked against before it is highlighted,
which files that check reads, whether the grab list is cleared on a CQ or a
QSY, what happens to AFC and the mark tone on a run to S&P change, and whether
the receive pane is copied to a text file.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
GrabIntoFocused puts text in whichever box has the cursor, whatever kind of
value that box normally holds. It is what N1MM's digital window does on
Ctrl+click, for the exchange element the contest has no box for.
The entry window also says when a CQ has gone out, which the digital window
needs to clear its grab list on: a run that starts again brings new callers.
GrabList.N1mmDepth is the ten calls N1MM's grab window holds.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
RTTY sends one shift character to say whether the codes that follow are
letters or figures. When it is lost to noise the printer stays in the wrong
case for the rest of the line, so 599 prints as TOO, 73 as UE and CQ TEST as
:1 53'5.
Baudot.LettersFigs swaps the case of a word using the US teleprinter table,
which is the one N1MM uses in DigitalMMTYIo.Letters_Figs.
N1MM converts a letter to its figure but not a figure back to its letter, so
its box turns VE4AEO into ;3R-39 and cannot turn it back. The table here is
read both ways, which is what MMTTY's own right click does: a call printed in
the wrong case is the case an operator has to read.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Something already sitting where Documents/N1MM Logger+ should go — a file, or
a link to somewhere that is not there — threw out of Directory.CreateDirectory
before the program had a window, so the operator got a stack trace and nothing
else.
CreateFolders answers with what stopped it, and the program shows it in a
window naming the path and stops there. NONEMM_HOME moves the whole lot
somewhere else, for a station that keeps its files off the Documents folder.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
The root is Documents/N1MM Logger+ now, so a station that runs both programs
has one set of files rather than two: one Databases folder, one SupportFiles,
one Wav with a folder per operator in it. What N1MM has no place for goes in a
Nonemm folder inside it, which is where settings.json sits.
A system that does not say where Documents is — most Linux systems — gets
~/Documents/N1MM Logger+.
Files an older version kept in Documents\Nonemm or ~/.config/nonemm are copied
in the first time this version runs, and the log and call history paths in the
settings are moved with them, so the copy that opens is the copy that came
over. The older folder is left as it was.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
The folders under the user's own directory are N1MM's now, with N1MM's names,
read out of N1MMLibrary.dll: Databases, SupportFiles, UserDefinedContests,
FunctionKeyMessages, CallHistoryFiles, ExportFiles, GoalFiles, LettersFiles,
QsoRecording, SkinsAndLayouts, SystemFiles, TransactionLogFiles, Diagnostics
and Wav. A station that runs both programs keeps its files in one shape.
An operator's recordings go in Wav/<his callsign>, which is where N1MM plays
them from, and the folder is made when he takes the radio. Cabrillo and ADIF
start in ExportFiles, call history files in CallHistoryFiles, and the .mc
function key files in FunctionKeyMessages.
N1MM's three *DDL folders and its Piper folders are left out: they hold the
migrations for N1MM's own admin databases and the program it speaks with.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Ctrl+O asks who is at the radio, as it does in N1MM, and the callsign goes on
every contact logged from then on. A multi-operator station changes hands every
few hours, so each operator also keeps what he wants the program to look like:
where the windows sit, the theme, and the folder his recordings are in.
The windows on the screen are stored against the operator leaving, before the
dialog opens, and the operator taking over gets his own back. A window he opens
later is placed as it appears. An operator who is new to the list starts with
whatever is on the screen.
Nothing plays the recordings yet, so the folder is stored and waits on voice
keying. {OPERATOR} in a message sends the callsign.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Window > Grey Line opens the world with the night on it. The dark area is
where the sun is down, and the band along its edge is the twilight from the
horizon to six degrees below it, which is the grey line an operator watches
for on the low bands.
Grayline works out where the sun stands: the subsolar point, the elevation at
a place, and the latitude the terminator crosses a meridian at, which is what
the map fills the night between. It is the same question SunTimes answers from
the other end, and the two are checked against each other in the tests.
The coastlines are Natural Earth's 110m coastline, public domain, cut to a
tenth of a degree.
Our own station is drawn from the grid square, the station being called and
the spots on the band from the country file. The map follows the clock.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
MMTTY and 2Tone have no socket or pipe interface: N1MM hosts XMMT.ocx and
exchanges window messages with the engine. A Linux process cannot load that
control, so bridge/nonemm-mmtty-bridge.exe hosts it under Wine and passes
lines over its standard input and output. docs/digital-bridge.md states the
protocol and what the control needs.
The window is N1MM's: receive pane with coloured callsigns, grab list, call
stacking, twenty-four macro buttons and the engine controls. One left click
copies what is under it — a callsign to the callsign box, anything else to the
exchange box the contest keeps for that kind of value.
Config > Digital registers XMMT.ocx in the Wine prefix on its own, making the
prefix first if it is not there. The engine, the bridge and the control start
at the copies shipped beside the program.
The bridge reads the control's own events. OnTranslateMessage carries only the
messages the control has no event for, so nothing was ever decoded through it.
hamlib's data mode names read as digital modes now, and a mode typed into the
callsign box changes mode the way a frequency changes band, so a station with
no radio can reach RTTY at all.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
A call history file is built from years of logs, so its section and
state columns carry old names, and the file says at the top what to do
about that. The four lines that say it were read and passed over.
All four are now carried out, in N1MM's order: !!MapOnSection!! stores
the old Ontario sections as ON, !!ValidateArrlSection!! drops a section
that is not one of N1MM's, !!GTA2GH_NT2TER!! stores GTA as GH and NT as
TER, and !!Validate50State!! keeps a state, turns a section into the
state it is in, or drops the value.
The section list is the one N1MM's call history reader validates
against, which is not the list its entry window offers: it holds the
retired GTA and NT, and leaves out ON. It sits beside the reader in
Nonemm.Core rather than with the contest lists, which are the entry
window's.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
SOBandChangeTimerDuration and MOBandChangeTimerDuration say how long a
station has to stay on a band it has just moved to. Both were read by
nothing: N1MM counts them down in its info window, which this program
does not have.
The countdown now stands in the entry window's status bar, next to the
contacts and multipliers, ticking with the clock that is already there.
BandChangeRules.StayLeft works it out from the log, starting the stay at
the minute after the contact that changed band, which is where N1MM
starts it.
Which contests show it is decided at runtime rather than by a list: the
rules come from the contest for the entry's own category, so a file that
asks for no stay, or asks for one only of single-op entries while the
station is entered multi-op, shows nothing at all.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
docs/unfinished.md said REF, ARRL 10M, the Ukrainian DX contest and the
Russian DX contest were not written yet. All four are in
src/Nonemm.Contests/Rules, and the first three implement SkipsField for
the country-dependent exchange the sentence is about.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Four settings were read by nothing.
MultipleSessions = 0000/30 runs the contest in half-hour sessions from
00:00 UTC. The session is part of the dupe key, so a station worked in
an earlier one may be worked again, and ResetMultsEverySession and
ResetSNEverySession start the multipliers and the serial numbers over.
DupeQSOMinutesAgo is the other way back to a station: after so many
minutes rather than in the next session. The log holds the time of the
last contact per dupe key instead of only the key, and the setting's
IgnoreBand and ThisMode decide the scope, because a file that uses it
turns DupeType off.
MinimumOffTime drives the time on and time off the score summary now
shows under the score, and beside them are the band changes made against
the allowance for a contest that limits them. N1MM's countdown for the
minutes you have to stay on a new band needs its info window and is not
here.
UdcFile.Flag now takes 1 and 0 as well as True and False, which is what
these settings are written as and what N1MM reads.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
MultSqlString = CallHist counts every station the call history file
lists, once per callsign, which is how N1MM dedups it. The file reaches
the scoring code through ContestLog.History, and a QsoContext now
carries it; without a file loaded nothing counts.
BonusPoints2 = +50, calls.txt reads the callsigns from the support-files
folder and adds 50 to the contact, *2 doubles it and a plain 50 scores
50 instead. The folder comes from the registry, which the app passes
when it reads the .udc files. N1MM's other form, where the file is
grids.txt and the bonus is looked up by grid square, is not read.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
MultWindowType names the list N1MM validates a received section against,
and every exchange that was not all digits counted as a multiplier here.
Four of N1MM's lists are held now -- the ARRL sections, the US states,
the Canadian provinces and the 165 OK/OM districts, which came out of
N1MM's own exe -- and a file naming one of them counts only the values
on it. A file naming any of the hundred-odd others keeps the old rule.
The 2023 OK/OM DX log now scores contact for contact as N1MM did. Two
contacts in the 2026 log still differ: the district code is stored with
a trailing space, which N1MM matches against its list untrimmed, so it
counts no multiplier where this counts one.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Alternating CQ ran off every message the keyer finished, so working a
station that answered started a CQ on the other radio, and Escape --
which stopped it outright -- was the way around that. It now follows
N1MM: only the CQ message carries the alternation to the other radio,
anything else holds it where it is, and so does Escape. The next CQ
picks it up again, which is what the CQ after a logged contact does, so
Ctrl+B is needed once rather than after every contact.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
The BARTG RTTY contest sends the time with the serial number, and both
stations log the time the other one sent. The entry window now has the
report, serial and time boxes N1MM's has, and the time this station sent
is stamped into the MiscText column as the exchange message goes out, or
at log time when nothing sent it. The Cabrillo line writes that time and
falls back to the time of the contact when the column is empty, which is
what N1MM writes, in N1MM's columns.
The total score was points times every multiplier. N1MM's class works
out points times the countries and areas together, times the continents.
Per-contact points and multiplier flags do not change, so the station's
2022 log still scores contact for contact as N1MM did.
N1MM has a class each for the March HF RTTY contest and the September
sprint. This program has one contest holding the March rules under the
sprint's name, and docs/unfinished.md now says so.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
CabrilloFormat 2, 3, 4 and 5 name the NAQP, NA Sprint, Sweepstakes and
section-and-serial lines. A file asking for one of those got the default
line; it now gets the layout, ported from CabrilloString2 and
CabrilloString4 in N1MM's Contact.cs, in N1MM's columns.
The exchange the layouts split into columns is built from the boxes the
file defines, in the order it defines them. N1MM builds it from the
section box alone, which writes only half of a two-box exchange such as
NAQP's name and state.
Format 6, the ARRL RTTY Roundup line, is still not read. No published
.udc file asks for it, because the Roundup has a contest class of its
own.
ExchangeSlots.ValueOf moves to Nonemm.Contests as QsoExchange.ValueOf,
so the contest code can read a contact through its exchange boxes
without a second copy of that switch.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
`CabrilloString` names a field and a column width in turn. The line it makes
follows N1MM: the operator's callsign in the first 13 columns, then every
column padded to its width with one space after it, and a value longer than its
column cut rather than pushed sideways. The cut is new for every contest, not
just the user-defined ones — the columns are the sponsor's, not ours.
`CabrilloVersion` is read too, so a contest whose sponsor still asks for 2.0
gets it in the header.
A contest that lays out its own line can put the transmitter column where it
wants it, or leave it out, so the writer no longer adds one at the end.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
I wrote SA 10 and OK/OM DX SSB in code after two failed attempts at the
download. The download works — the page posts a nonce with the form — and the
files say I got both contests wrong.
SA 10 is deleted from the code. Its file scores the station's log to the
contact, where the code had the wrong Cabrillo name and counted its multipliers
once in the contest rather than once per mode.
OK/OM DX now covers the CW running only, which is what N1MM's class is for and
what N1MM's own documentation says: the SSB running has different rules and a
pair of .udc files, one for each side. The CW rules go back to the class, so
Czechia and Slovakia are one side of the contest again. With the SSB files in
place both of the station's SSB logs score to the contact.
Reading those files turned up four things in the .udc reader: a file whose
extension is upper case was passed over, `Country` was not read as a
multiplier source, `OtherCountry` was not a points condition, and a section
multiplier counted the serial numbers the other side of a contest sends.
All Asian stays in code — N1MM has a class for it after all — with the Cabrillo
name and the band table it uses.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
REF, the Ukrainian DX contest, the BARTG Sprint, ARRL 10M, Russian DX RTTY,
RDAC, YO DX HF, Oceania DX, IARU Region 1 Field Day, SARTG RTTY, All Asian and
SA 10. Every contest name in the log now opens, and every one of these but
Field Day scores its log contact for contact as N1MM did.
Most are written from N1MM's own class and checked against the log. Three
things the log settled that the class does not say plainly: ARRL 10M counts its
multipliers once per mode rather than once, the BARTG Sprint counts a continent
once in the contest while its countries and call areas count per band, and RDAC
counts nothing but the districts and scores nothing for a station that sends
none.
All Asian and SA 10 have no class in N1MM, which logs them from a .udc file, so
those two are written from the published rules and the log. docs/unfinished.md
lists what is still not settled.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Four more of the contests in the station's log. IOTA, CQ 160 and Mexico RTTY
score every contact in it exactly as N1MM did.
IOTA pays 15 points for a station that sends an island reference and 2 for one
that does not, and counts each reference once per band and mode. CQ 160 is one
band, so its states and countries count once each. Mexico RTTY and CQ 160 both
take a state from one side of the contest and a serial or a zone from the
other, in the columns N1MM keeps them in.
The EU DX Contest is written from N1MM's class, including its list of what
counts as Europe: the European Union, its outermost regions and its overseas
territories, and nothing else. The station's own logs were made with the .udc
file, which scores and counts two things differently; docs/unfinished.md says
which.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Both are in the station's log and neither could be opened. The rules are
written from the published ones and checked by replaying the log.
YOTA scores a contact by the age the other operator sends: 13 points down to 10
for anyone 25 or under, whatever the continent, and 1 point or 3 for everyone
else. Every age worked on a band is a multiplier. The 2024, 2025 and 2026 logs
come out to the contact.
OK/OM DX is the first contest whose exchange differs by the other station's
country: a county from OK and OM stations, a serial number from everyone else,
in the two columns N1MM keeps them in, and `SkipsField` walks the entry window
past the box that does not apply. Points depend on which side the operator is
on, and Czechia and Slovakia count as the two countries they are. The 2023 log
comes out to the contact.
The Cabrillo exchange now gets the entry, because YOTA sends an age that only
the contest setup knows.
Two differences are left in the log, both from the .udc files the operator
used: their YOTA file only lists ages 7 to 25, so a younger operator scored as
an adult, and one OK/OM log was made with the file for stations outside OK and
OM while signing an OM call.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
The three differences left after replaying the log were ours on purpose. A log
has to score the same in both programs, so they now follow N1MM:
- ARRL DX and CQ WW RTTY count the 14 Canadian areas N1MM lists as `48SDC14P`,
which keeps Newfoundland apart from Labrador. A station sending the postal
code is counted as the area the contest names.
- A maritime mobile is scored from the call it signs. CQ WW says it counts for
the zone alone; N1MM gives it the country, and so do we.
- A callsign the country file cannot place counts as a multiplier with an empty
value, so every unplaceable call worked on a band shares one multiplier.
29 of the 37 contest instances in the log now agree with N1MM on every contact
and on the totals. What is left is the country file of the day and rows whose
cached points N1MM never worked out again after a call was corrected.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Two pieces of work on the scoring engine.
The .udc reader now takes N1MM's published scoring vocabulary: the full
PointsPerContact condition set, the PointsMultBy family, PowerMult, BonusPoints,
the multiplier sources and the settings that say which stations bring a
multiplier in. MultMult is a weight rather than a switch, as in N1MM's
ComputeScore. The entry categories and the sent exchange reach the rules in a
new ContestEntry, so a contest can score by the power category it is entered in
or by what this station sends.
Then every contest in a real N1MM log was scored again from these rules and
compared with the points and multiplier flags N1MM wrote. That found five bugs:
ARRL DX and IARU read the exchange from the wrong column, a stored " " was read
as a value rather than as empty, CQ WW RTTY's own Canadian area names were not
counted, a maritime mobile scored nothing, and a contact in a mode the contest
does not run scored as if it were in the contest.
docs/n1mm-interop.md has the check and what still differs.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Reading the source again after this was tried turned up several things it had
wrong or left out.
The TU message went out only when reading a series out. N1MM sends it whichever
way the traffic went, and lets it name the series: {QTC} in it stands for the
header, so TU {QTC} 73 goes out as TU QTC 3/10 73.
The RX Ready button was hidden while sending and sent nothing on CW. It is
N1MM's QRV button: reading out it says R U QRV and keys QRV?, taking down it
says QRV and keys QRV.
The fields were asked for again with shift and Enter, which was made up. N1MM
uses the number keys, and writes them under the lines: 1, 2, 3 and 4 send the
time, call, serial or header of the last line again while reading out, and
shift with 1, 2 or 3 asks for that field while taking down.
The buttons now carry N1MM's names for the direction the traffic is going —
R U QRV, Snd Hdr, Snd1 to Snd10, Exit reading out; QRV, Hdr Agn, Agn1 to Agn10,
Exit taking down — and the Cfm column and Clear are hidden while reading out,
as N1MM hides them. The separate Send Hdr button is gone: N1MM has one button
there that sends the header one way and asks for it the other.
The log window gained N1MM's QTC? column for a contest that carries QTC
traffic. It holds the series, so a QTC row can be told from a contact; before
this a QTC row read as an ordinary one with odd values in the report columns.
Driven under Xvfb against a fake cwdaemon. Reading out: QRV?, QTC 3/2,
2319 DL7GGG 03, then 2319 and QTC 3/2 from the number keys, and
TU QTC 3/2 73 on Exit. Taking down: QRV, AGN TIME, AGN CALL.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
The QTC window took traffic down and read it out of the log, but nothing left
the window: on CW the operator had to send the lines by hand. It now puts them
on the air through the entry window's keyer, the way N1MM's QTC window sends
through its own entry window, so the same macros are expanded and the same
transmitter is pointed at first.
Sending a series: Send Hdr keys the header, Send1 to Send10 key one line each —
the time, the callsign and the serial number — and Close keys the TU message
when one is set. Taking one down: Hdr Agn asks for the header again, and shift
with Enter in a box asks for that one field again and clears it.
The messages and their defaults are N1MM's: AGN TIME, AGN CALL, AGN SERIAL, the
hard-coded HDR AGN with its two spaces, and a TU that sends nothing until the
operator writes one. So is the field spacing, written N1MM's way with S for a
space. All five are in Setup, where the note about the missing settings used to
be.
Nothing is sent on SSB or RTTY. N1MM plays four recordings on SSB and sends
RTTY from its digital window, and there is neither here, so on those modes the
buttons move the cursor as before. RX Ready sends nothing on any mode, because
a recording is all N1MM ever sends from it. Serial numbers go out as digits:
N1MM can send cut numbers and nothing here does, in any message.
Driven under Xvfb against a fake cwdaemon, in a WAE CW contest with four
contacts logged. Sending, it keyed QTC 1/2, 2257 DL1AAA 01, 2257 DL2BBB 01,
then QTC 2/2 with its two lines and TU DE W1ABC on close. Receiving, HDR AGN,
AGN TIME, AGN CALL and AGN SERIAL.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Three parts, as N1MM has them: what just happened on the left, the contacts
and multipliers in the middle, the score on the right.
The left says where the call being typed is from, in N1MM's own shape:
OM: EU -> Slovak Republic, Zn 15. The zone is the one the contest counts in, so
Contest.UsesItuZones answers that the way N1MM's ZoneType does, and IARU
overrides it. A message from something the operator just did stands there until
the next keystroke.
The middle is N1MM's breakdown: the contacts and then one number per multiplier
the contest counts, separated by slashes, and No Score before the first
contact. Points are not in it, because they are not in N1MM's; the score window
has them.
Those numbers were in the line at the top of the window, which now carries the
contest name alone, and the country was in the bar under the callsign box,
which now says only what working the call would bring. Neither is said twice.
Looked at under Xvfb: the country line for OM3KFF and then G4AB, 10/7/8 and
300 beside them, and a Mark message standing in the left until the next
keystroke.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
AppSession.TakeFromNetwork wrote the contact to the store and then called
OpenContest to see it, which built a new ContestSession, read the whole log
back and built new entry positions. Anything the operator was typing at either
radio went with them, and the entry boxes were rebuilt under the other
operator's hands. It also ran on the socket thread, so it was changing the log
while the windows were reading it.
ContestSession now applies the three messages itself: AddFromNetwork,
ReplaceFromNetwork and DeleteFromNetwork write the store, change the log in
place and raise Changed. None of them raises Logged, Edited or Deleted, so
nothing goes back out to the other stations. Points and multipliers are still
worked out here from the rules rather than trusted.
A contact that arrives goes in where its timestamp says it belongs, through
the new ContestLog.Insert, and the log is scored again: of two stations that
worked the same multiplier, the one that worked it first keeps it, however
late the message turns up. Add would have appended it at the end and given the
multiplier to the wrong contact.
The socket thread now hands the update to the thread the windows run on, which
is where every other change to the log is made.
The matching rule moved with the code, so ContestSession answers what a
message from N1MM refers to: the contact id, then the call and time it had
before the edit.
Still not tested against a second station. Two of these cannot run on one host
to try it, because they share one UDP port.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Both put something on our own bandmap and nothing on the cluster, which is
what separates them from Spot It.
Mark, Alt+M, leaves a mark where the radio is to say the frequency is busy.
The label is N1MM's, Busy@ and the time. It is not a station, so Spot carries
IsStation: nothing judges it, the check window does not offer it as a call,
available mults and Qs leaves it out, space does not take it into the callsign
box, and the bandmap paints it like a station already worked, which is how
N1MM marks it. The station count under the bandmap counts stations.
Store, Alt+O, puts the call being typed on the bandmap at this frequency so it
can be worked later. N1MM writes "Local spot" in the comment and this does the
same.
The action row now holds the eight buttons N1MM has, in N1MM's order and at
N1MM's widths: 23 per cent to Esc: Stop and 11 to each of the rest.
Left out of Store: N1MM also adds the call to what the check window's Master
column offers for the rest of the session. The call is on the bandmap, which
the check window already reads, so it is offered either way.
Driven under Xvfb: Alt+M leaving BUSY@22:32:27 on 14000, Alt+O storing
SP9XYZ there, and both drawn on the bandmap in the worked colour.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
A CQ is answered by more than one station. N1MM lets the operator put the
others on a stack and work them one after another; this is that, with N1MM's
keys and macros.
Ctrl+Alt+G and {STACKANOTHER} put the call being typed on the stack and clear
the boxes. Alt+G brings the top call back. Alt+D drops it. {SOCALLSTACK} both
stacks and unstacks in one key: it swaps the typed call for the top one, stacks
it when the stack is empty, and takes the top call when the boxes are.
{LOGTHENPOP} and {LOGTHENNEXT} log the contact and bring the next caller in.
{CLRSTACK} empties it.
The rules are N1MM's, and CallStack and the entry actions hold them without a
window, so they are unit-tested: stacking only while running, nothing shorter
than two characters, no dupe, no frequency typed into the callsign box, a call
already on the stack moved rather than repeated, and a multiplier to the front.
{SOCALLSTACK} takes the stacked call off before it puts the typed one on, so a
multiplier just typed cannot come straight back.
Two places differ. N1MM refuses your own call only while the stack is empty;
it is never a station to work, so it is refused either way here. And the order
is multipliers first in every mode: N1MM's digital window offers first-in and
last-in as well, and there is no digital window here.
The stack window opens by itself with the first call and closes with the last,
or Window > Call Stack opens it. It does not take the keyboard when it opens,
because the operator is typing the next call. The next call out is ringed and
each is coloured by what working it would bring, as in every other window.
OperatingPosition.JudgeCall now answers what a call alone is worth, which the
stack window and the check window both need; it was private to the check
window before.
Driven under Xvfb: four calls stacked with Ctrl+Alt+G and listed in the window,
Alt+G bringing the top one into the boxes, Alt+D dropping the next.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
The twelve function keys were one row of twelve at the bottom of the window,
with the action row under them. N1MM puts both between the entry boxes and the
bearing line, six keys to a row in two rows, each row filling the width.
The action row takes N1MM's widths as well: 23 per cent to the first button and
the rest shared out. N1MM gives the remaining seven 11 per cent each; there are
five here, because Mark and Store are missing, so they share the same 77.
Looked at under Xvfb against the N1MM screenshot: same two rows of six, same
action row under them, same order down the window.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
N1MM's band panel is a row per band and a column per mode the contest runs,
with every cell painted by what the callsign being typed would be worth there:
red for one new multiplier, green for more than one, blue for points, grey for
a station already worked on that band and mode. With nothing typed every cell
is grey. That is the panel's point — it says where a station is still needed
without working the bands one at a time — and this had only the frame.
BandPanel holds the rule and is unit-tested without a window. The bands are the
six contest bands, with 30, 17 and 12 metres added for general logging: in N1MM
that is the one entry of the hundred and thirty-five that shows the WARC bands,
so Contest.ShowsWarcBands says so and GeneralLogging overrides it. The columns
are the modes the contest runs, so CQ WW CW has one column and IARU has two,
rather than always showing CW and phone. A digital entry gets one DIG column
where N1MM has RTTY and PSK: dupes and points here go by mode category, and
those two are the same category.
The labels are N1MM's: the metres alone up to 10, then the unit from 6m up. The
band the radio is on is ringed rather than filled, as N1MM rings it, so the
ring does not hide the colour underneath. The cells are labels with a click
handler rather than buttons, so the theme cannot paint over the colour under
the pointer.
Text on a filled cell is black or white by N1MM's rule: light is the three
channels adding up to more than 384.
Looked at under Xvfb in CQ WW CW: one CW column, six bands, grey with nothing
typed, and with a JA typed red on every band but 20, where the zone and country
are already claimed, and which carries the ring.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
N1MM's log window has two panes. The lower one lists every contact already
logged with the call in the callsign box, on any band and in any mode, so an
apparent dupe can be read off the screen rather than looked for.
EarlierContacts.Matching holds the rule, so it is unit-tested without a
window: three characters before anything is listed, a call that holds what is
typed, `*` for any run of characters and `?` for one, the grid square instead
when nothing is typed, ordered by call, mode, band and time, and fifty at
most. That is N1MM's DupeSQL, except that N1MM passes `?` to SQLite as a plain
character, which finds nothing.
The two grids share their columns, because a value in the pane has to sit
under the column it belongs to. Each column is now as wide as the longest
value it can hold, measured in the font it is drawn in, and the last column
takes what is left over. The values and the header are in different fonts, so
each is measured on its own. This also settles the column widths, which used
to change as rows scrolled into view.
A DataGrid does not pass its own font down to its cells, so the cell font is
set in a style. Without that the cells draw larger than the width measured for
them and every column is a character short.
Looked at under Xvfb with a call typed: the pane lists the two contacts that
hold it, under columns that line up with the log above.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01RoGtneMQaz4M9w7Kk49AVD
Colours came from static brushes scattered over the windows. They now come
from a Theme, and everything that paints a station, a spot or a log row
asks Themes.Current as it paints, so changing the theme changes every
window without reopening it.
light is N1MM's own palette, read out of ApplicationStyles: the pale blue
behind a window, ivory on a button, white boxes, and the four colours a
station is painted in. dark is the same states in colours for a dark
screen, which is roughly what the program looked like before.
Config > Manage Skins, Colors and Fonts picks one, under N1MM's name for
the same dialog, and shows what each theme paints a callsign. The choice
is kept in settings.json as Theme, and light is the default.
The window chrome follows through resources the Fluent controls read, so a
text box and a button take the theme's colours in every state. Fonts and
skins are not part of it.
Looked at under Xvfb in both themes: the entry window, the theme dialog,
and the log window with its rows still coloured by what each contact was
worth.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The colours are N1MM's own defaults, read out of ApplicationStyles: blue
for a station worth working, red for one new multiplier, green for more
than one, grey for a dupe, and red on the CQ frequency line. A call under
three characters stays blue, as N1MM leaves it.
The mark at the right-hand end of the callsign box is N1MM's OKCall: a
tick when the call is in MASTER.SCP, a question mark while it is not, in
the colour of the call. Like N1MM it only shows while the check window is
open, for three characters or more, and only while the call is short
enough to leave room for it.
The band under the boxes takes N1MM's background colours instead of white
on the text colours. The two lights, the telnet window and the message
editor were borrowing the verdict brushes for things that are not
verdicts; they have their own colours now.
The callsign and the exchange boxes uppercase what is typed. Name and
comment do not, which is what N1MM does.
Scored contest 63 of om5m.s3db (CQ WW SSB, 3775 QSOs) against the points
and multiplier flags N1MM stored: points identical, multipliers 601
against 602. The difference is B1Z on 10m, which N1MM logged with an
empty country prefix and counted as a country of its own though BY was
already worked on that band.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
